Re: Bias-exchange metadynamics

From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Sun Nov 19 2017 - 17:30:49 CST

Hi Azadeh, an easy-to-use implementation of bias-exchange metadynamics with
NAMD+Colvars has not been traditionally a high priority. Because the
number of replicas with BE-MTD can quickly get very high, obtaining a large
enough allocation of nodes for NAMD is difficult. It has often been
efficient to use few nodes per replica, and using MD programs that run fast
at lower processor counts, such as GROMACS. PLUMED provides an
implementation of BE metadynamics that is well integrated with GROMACS,
with the caveat that you need to recompile the patched code on your own.

More recently, the paradigm that NAMD is better suited for fast
single-replica jobs over large sets of nodes is starting to change. For
example, GROMACS has become more parallel and GPU-based NAMD jobs are fast
on recent hardware. But this a gradual process and as I mentioned, the
priority is not on replicating functionality that is well implemented in
other codes.

That said, last January I answered a question about BE-MTD to you on this
mailing list:
http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2017-2018/0019.html
where I provided you with examples of commands to add to the existing Tcl
replica scripts, specifically replacing bias-exchange umbrella sampling.
Have you made some progress? Where did you stop?

If you got stuck, here is my suggestion:
1) Make a copy of the umbrella.namd script for you to modify freely;
2) In that file, locate the lines that implement the bias-exchange umbrella
sampling, identify all NAMD commands in them (i.e. not Tcl keywords) and
study their documentation;
3) If there are any problems with the previous step, please report back to
this list. The replica-exchange scripts should be readable by somebody
with good knowledge of Tcl: if they are not, this should be fixed, starting
with comments that clarify what the various steps are.

Giacomo

On Fri, Nov 17, 2017 at 6:07 AM, Azadeh Alavizargar <
azadeh.alavizargar_at_gmail.com> wrote:

> Hello everyone
>
> I am going to run bias-exchange metadynamics with NAMD. I only know that I
> should use TCL scripting of NAMD and the colvars scripting tool. However,
> it sounds complicated to me.
>
> Do I need to modify only the replica.namd script in the lib/replica
> directory of NAMD? I suppose I should omit the temperature exchange part.
> Am I right? What else do I need to remove or add?
>
> What about the colvars scripting according to the link below for colvars
> and biases?
>
> http://colvars.github.io/colvars-refman-namd/colvars-refman-
> namd.html#sec:scripting
>
> I think I need to have multiple colvars files and load and unload them
> through "cv" command. But I do not know where exactly and how I should use
> them.
>
> As far as I know, there are no examples on how to use these capabilities
> and set up a bias-exchange simulations, especially for the colvars
> scripting part. I would be thankful if anyone can help me on this issue.
>
> I really appreciate your help.
>
> Best regards,
> Azadeh
>

-- 
Giacomo Fiorin
Associate Professor of Research, Temple University, Philadelphia, PA
Contractor, National Institutes of Health, Bethesda, MD
http://goo.gl/Q3TBQU
https://github.com/giacomofiorin

This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:47 CST